Search for Gravitational Wave Transients Florent Robinet On behalf - - PowerPoint PPT Presentation

▶
search for gravitational wave transients
SMART_READER_LITE
LIVE PREVIEW

Search for Gravitational Wave Transients Florent Robinet On behalf - - PowerPoint PPT Presentation

Sep 15, 2022 190 likes •533 views

Search for Gravitational Wave Transients Florent Robinet On behalf of the LSC and Virgo Collaborations Rencontres de Moriond - March 2011 1 Gravitational Waves Gravitational Waves Gravitational waves = "ripples" in space-time

slide-1
SLIDE 1

Search for Gravitational Wave Transients

Florent Robinet On behalf of the LSC and Virgo Collaborations

Rencontres de Moriond - March 2011 1

slide-2
SLIDE 2

Gravitational Waves Gravitational Waves

Black hole merger

Gravitational waves Gravitational waves = "ripples" in space-time = "ripples" in space-time Weak field approximation : Weak field approximation : Wave equation, speed Wave equation, speed c c Solution with 2 d.o.f. : Solution with 2 d.o.f. : Dimensionless amplitude given by Dimensionless amplitude given by h h Signal strength: Signal strength: Production of gravitational waves A good GW source :

  • is compact and massive
  • is asymmetric
  • has a relativistic speed

c

5

G 

2

Rs R 

2

v c 

6

ℒ =

Lab production : Lab production : h ~ 10 h ~ 10 – 39

– 39

Astrophysical sources : Astrophysical sources : h ~ 10 h ~ 10 – 21

– 21

g=h ∣h∣≪1

Rencontres de Moriond - March 2011 2 Florent Robinet

h=h+hx

hrss=∫-∞

+∞

∣h+t ∣

2∣hxt∣ 2dt

slide-3
SLIDE 3

Gravitational Wave Sources Gravitational Wave Sources

Rencontres de Moriond - March 2011 3 Florent Robinet

?

Pulsars (asymmetric rotations, instabilities) Compact binary coalescence of neutron stars &/or black holes Supernovae (asymmetric core bounce) Cosmic strings Stochastic background The unexpected

slide-4
SLIDE 4

A Network of Detectors A Network of Detectors

Virgo (3 km) Geo (600 m) Livingston (4 km) Hanford (4&2 km)

LSC – Virgo collaboration

– Full data sharing since May 2007 – Common analyses and papers – Common tools Rencontres de Moriond - March 2011 4 Florent Robinet

slide-5
SLIDE 5

A Network of Detectors A Network of Detectors

tLivingston tHanford tHanford tVirgo SOURCE SOURCE POINTING

  • Source location within ~ tens of square degrees
  • Serious candidates follow-up (EM, neutrinos...)

GHOST Rencontres de Moriond - March 2011 5 Florent Robinet

slide-6
SLIDE 6

A Network of Detectors A Network of Detectors

Rencontres de Moriond - March 2011 6 Florent Robinet

Hanford sky coverage – Antenna pattern

slide-7
SLIDE 7

A Network of Detectors A Network of Detectors

Rencontres de Moriond - March 2011 7 Florent Robinet

Livingstone sky coverage – Antenna pattern

slide-8
SLIDE 8

A Network of Detectors A Network of Detectors

Rencontres de Moriond - March 2011 8 Florent Robinet

Virgo sky coverage – Antenna pattern

slide-9
SLIDE 9

A Network of Detectors A Network of Detectors

Rencontres de Moriond - March 2011 9 Florent Robinet

Network sky coverage – Antenna pattern

slide-10
SLIDE 10

LIGO / Virgo Science Runs LIGO / Virgo Science Runs

Rencontres de Moriond - March 2011 10 Florent Robinet 2005 2006 2007 2008 2009 2010 2011 2012

S4 S5 S6 VSR2 VSR1 VSR3 VSR4?

Advanced Detectors Many Publications Analyses in progress Publications in preparation commissioning

slide-11
SLIDE 11

LIGO / Virgo Science Runs LIGO / Virgo Science Runs

Rencontres de Moriond - March 2011 11 Florent Robinet

S5 - VSR1 S6 - VSR2

slide-12
SLIDE 12

Analysis Groups Analysis Groups

Compact Binary Coalescence (CBC) Short Signals (Bursts) Continuous Waves Stochastic

?

Rencontres de Moriond - March 2011 12 Florent Robinet See C. Palomba's talk

slide-13
SLIDE 13

Analysis Groups Analysis Groups

Compact Binary Coalescence (CBC) Short Signals (Bursts)

?

All sky GRB-triggered SGR Flares EM Follow-up Binary Mergers Pulsars Glitches Supernovae Multi-Messenger Astronomy Low Mass High Mass Inspiral-Merger- Ringdown (IMR) Parameter Estimation Cosmic Strings

Rencontres de Moriond - March 2011 Florent Robinet

slide-14
SLIDE 14

CBC vs. Bursts CBC vs. Bursts

Rencontres de Moriond - March 2011 14 Florent Robinet Compact Binary Coalescence (CBC) The Signals

  • Modeled signals
  • Inspiral – Merger – Ringdown

The search

  • Template search (selective)
  • Waveform parameter estimation

Science goals

  • Detection
  • Upper limits on GW emission
  • Multi-messenger (EM, neutrino...)
  • Parameter estimation
  • Study gravity
  • Study populations
  • Study dense matter
  • Study GRBs

Burst Signals The Signals

  • Short-duration signals (<1s)
  • Modeled and unmodeled signals
  • Large variety of sources

The search

  • Robust signal detection methods
  • Excess power (unmodeled)
  • Template search (modeled)

Science goals

  • Detection
  • Upper limits on GW emission
  • Multi-messenger (EM, neutrino...)
  • Parameter estimation
  • Star equation of state
  • Study populations
slide-15
SLIDE 15

Analysis Methods Analysis Methods

Rencontres de Moriond - March 2011 15 Florent Robinet

Data Detector 1 Data Detector 2 Triggers Triggers Coincidence Selection + Data Quality Significance? Data stream

COINCIDENT PIPELINE

slide-16
SLIDE 16

Analysis Methods Analysis Methods

Rencontres de Moriond - March 2011 16 Florent Robinet

Data Detector 1 Data Detector 2 Triggers Triggers Coincidence Selection + Data Quality Significance wrt background Data Detector 2 time-shifted wrt 1 Data stream Background stream

DETECTION? NO? → Upper limits

COINCIDENT PIPELINE

slide-17
SLIDE 17

Analysis Methods Analysis Methods

Rencontres de Moriond - March 2011 17 Florent Robinet

Data Detector 1 Data Detector 2 Triggers Triggers Coincidence Selection + Data Quality Significance wrt background Signal injections Data Detector 2 time-shifted wrt 1 Search efficiency Data stream Injection stream Background stream

DETECTION? UPPER LIMITS

The number of detectors can be increased (up to 4) Various coincidence schemes: union of configurations Increasing the number of coincidences enables to be more selective (but less efficient)

COINCIDENT PIPELINE

slide-18
SLIDE 18

Analysis Methods Analysis Methods

Rencontres de Moriond - March 2011 18 Florent Robinet

Data Detector 1 Data Detector 2 Triggers Coherent combination Selection + Data Quality Significance wrt background Signal injections Data Detector 2 time-shifted wrt 1 Search efficiency Data stream Injection stream Background stream

DETECTION? UPPER LIMITS

The data of multiple detectors can be combined coherently Sky positions are scanned to take into account the time of arrival and the antenna pattern of each detectors

COHERENT PIPELINE

slide-19
SLIDE 19

Analysis Methods: Template Searches Analysis Methods: Template Searches

Rencontres de Moriond - March 2011 19 Florent Robinet Detector whitened Strain h(t)

Noise + inspiral hardware injection

A template bank covering the parameter space is slid over the data

t∝∫

∞ hf f 

Shf  e

−2i f t df

Signal-to-noise ratio time serie: Template waveform Sensitivity An event is defined when Given by the background estimation (+ additional clustering in time and frequency)

tThreshold

This method is used for:

  • CBC searches
  • Pulsar ringdown search
  • Cosmic string burst search

 Sh

slide-20
SLIDE 20

Analysis Methods: Excess Power Searches Analysis Methods: Excess Power Searches

Rencontres de Moriond - March 2011 20 Florent Robinet Injected Inspiral Signal The time-frequency plane is tiled with pixels An event is defined when the energy

  • f multiple pixels exceeds a given

threshold This method is used for:

  • Most of the burst searches
slide-21
SLIDE 21

Data Quality Data Quality

Rencontres de Moriond - March 2011 21 Florent Robinet

The noise of the detector displays a non- Gaussian behavior Transient glitches removal is crucial to improve the sensitivity of the searches Noise understanding for each detector have been performed for each science run Many glitch families have been understood Specific vetoes based on auxiliary channels have been produced to remove specific glitch families Veto safety have been carefully checked (we don't want to flag real signals!)

Example: Virgo, VSR2 Deadtime ~ 10% Removal efficiency ~ 80% (SNR>8)

SNR

10 100 Magnetic sensor Detection channel

slide-22
SLIDE 22

CBC Searches CBC Searches

Rencontres de Moriond - March 2011 22 Florent Robinet "Realistic" observable BNS coalescence rate ~ 0.02 per year (large uncertainty) BNS BBH BHNS S5/VSR1 data have been analyzed and results are published S6/VSR2-3 analyses are in progress Preliminary results are released See T. Dent's talk

slide-23
SLIDE 23

CBC: Low Mass Search CBC: Low Mass Search

Rencontres de Moriond - March 2011 23 Florent Robinet

Description of the search:

  • Non-spinning templates
  • Post-Newtonian up to the innermost stable
  • rbit
  • Mass region 2 < Mtotal < 35 Msun

BNS/BBH Upper limits NS/BH Upper limits No Detection

  • Phys. Rev. D 82(2010) 102001
slide-24
SLIDE 24

CBC: High Mass Searches CBC: High Mass Searches

Rencontres de Moriond - March 2011 24 Florent Robinet High-Mass

  • Effective One Body (EOB) waveforms
  • Inspiral-Merger-Ringdown is covered
  • No spin
  • Mass region 25 < Mtotal < 100 Msun
  • Uncertainty on the waveforms
  • LIGO only search (Virgo is not sensitive

enough for High mass systems) Ringdown search

  • In progress for S5
  • LIGO only search
  • Mass region

75 < MBH < 750 Msun

  • Spin is included
  • The ringdown contains

most of the GW energy

  • More reliable waveforms

arXiv:1102.3781

slide-25
SLIDE 25

CBC & Burst Signals: GRB Triggered Searches CBC & Burst Signals: GRB Triggered Searches

Rencontres de Moriond - March 2011 25 Florent Robinet Powerful bursts of highly energetic gamma rays Two populations: short and long duration Short: possibly produced by the merging of binary objects → CBC colored search Long: possibly produced by violent stellar collapse (hypernovae) Short-duration GRBs Long-duration GRBs Use mainly Swift and Fermi triggers to get a source location a timing and sometimes a distance Background reduction Better sensitivity During S5/VSR1 137 GRBs were analyzed by the burst coherent pipeline (short and long). See M. Was's talk 22 GRBs were analyzed by a CBC

  • pipeline. See N. Christensen's talk

20

Number of Bursts

40 60 0.1 1 10 T90 (seconds)

slide-26
SLIDE 26

CBC & Burst Signals: GRB Triggered Searches CBC & Burst Signals: GRB Triggered Searches

Rencontres de Moriond - March 2011 26 Florent Robinet

GRB 070201

Short and hard GRB detected by 4 satellites in Feb. 2007 in direction of the Andromeda galaxy

The binary merger scenario is excluded with a 99% confidence level!

90% 75% 50% 25%

Inspiral Exclusion Zone

99%

  • Astrophys. J. 681(2008) 1419
slide-27
SLIDE 27

Burst Signals: All-Sky Search Burst Signals: All-Sky Search

Rencontres de Moriond - March 2011 27 Florent Robinet

Description of the search

  • Multiple Algorithms
  • Broad-band frequency search
  • Coincident and coherent searches
  • Very large variety of waveforms
  • Robustness

Upper limit (sine-gaussian) Rate (90% C.L.) vs. frequency

No detection With a 90% confidence level, the rate of burst signals with 50 < f < 2048 Hz is lower than 2 events per year

  • Phys. Rev. D 81(2010) 102001

~8e-7 yr-1 Mpc-3 1e-2 yr-1 Mpc-3

100 f(Hz) 1000

(EGW = Msunc2)

slide-28
SLIDE 28

Burst Signals: Neutron Stars Burst Signals: Neutron Stars

Rencontres de Moriond - March 2011 28 Florent Robinet 2006/08/12: timing glitch observed in the radio emission of the Vela pulsar Quasi-normal mode oscillations GW Both Hanford detectors were up at that time Ringdown signal is searched → No detection Upper limits on GW energy released by mono- harmonic modes

  • Phys. Rev. D83(2011) 042001

Soft Gamma Repeaters (SGR) / Anomalous X-ray Pulsars (AXP) Neutron stars powered by extreme magnetic fields (magnetars) 6 magnetars have been analyzed by a dedicated pipeline (excess power)

SGR 1900+14 SGR 0418+5729 SGR 1627-41 SGR 1806-20 SGR 0501+4516 AXP 1E 1547.0-4508

arXiv:1011.4079

46 52

slide-29
SLIDE 29

GW GW

Burst Signals: Cosmic String Cusps Burst Signals: Cosmic String Cusps

Rencontres de Moriond - March 2011 30 Florent Robinet When 2 cosmic string segments meet they can reconnect and produce loops. The main mechanism for the loop to loose its energy is to radiate gravitationally.

GW radiation is the most promising signature to detect cosmic strings.

Points of the string can acquire a large Lorentz boost and form a "cusp" → GW burst S4 S5 Projection Well-modeled signal Template burst search Dedicated pipeline Upper limits on the cosmic string parameter space: Gμ: String tension

ε: loop size parameter p: reconnection probability

Gμ

Phys Rev D 80(2009) 062002

10 –6 10 –7

slide-30
SLIDE 30

Online Searches Online Searches

Rencontres de Moriond - March 2011 Florent Robinet Low latency searches took place during S6 / VSR2-3 for both CBC and burst searches Most significant events were sent to telescopes / satellites (14 events for S6/VSR2-3) Sky localization is performed with a resolution of ~tens of square degrees for events at threshold Image analysis is performed within the collaboration See M. Branchesi's talk

Hanford Livingstone Virgo Central Location CBC pipeline MBTA Burst pipeline Omega Burst pipeline

Coherent Waveburst

Candidate Database Most significant Candidate selection

h(t) h(t) h(t) h(t) h(t) h(t) event event event

~ 30 min

Swift Zadko ROTSE

slide-31
SLIDE 31

S6/VSR2-3 Science Run S6/VSR2-3 Science Run

Rencontres de Moriond - March 2011 31 Florent Robinet

Improved sensitivity ~ 200 days of live-time with at least 2 detectors up Big challenge: run analyses online – 3 pipelines were running (2 bursts + 1 CBC) – The data quality was performed with low latency (< 1 min) – EM follow-up by partner telescopes Offline analyses are in progress (some results are released) Analyses pipelines and data quality tools have been improved for a better sensitivity Blind hardware injection challenge was successful (we detected it with great confidence). See T. Dent's talk

slide-32
SLIDE 32

Preparation for Advanced Detectors Era Preparation for Advanced Detectors Era

Rencontres de Moriond - March 2011 32 Florent Robinet Sensitivity improvement by a factor 10 This translates into a detection rate up to 40 neutron star binary events per year Science should resume in 2015 Design sensitivity achieved by 2019

What are we going to do in the meantime ?

Virgo might run this summer (VSR4) along with the GEO detector Similar sensitivity at high frequency This run could be of some interest for external triggered searches Then GEO will run alone in astrowatch mode during the construction of Adv. detectors (2012-2015) Some mock data runs are planned to test and improve

  • ur searches

A lot of work is required to optimize the EM-followup procedures

slide-33
SLIDE 33

Summary - Conclusions Summary - Conclusions

Rencontres de Moriond - March 2011 33 Florent Robinet

A large variety of physical results have been produced from the LIGO-Virgo data There is no detection yet but upper limits can be used to constrain astrophysical models The first generation of detectors is close to the end. – Analysis pipeline have greatly improved over the last years to perform

  • ptimized and sensitive searches on GW data

– Data quality is a great challenge. Multiple tools have been developed to reject noise events efficiently. – GW astronomy is on its way: online searches, Multi-messengers, EM followup Now, the big challenge is to be fully ready for the Advanced Detectors Era